Hi! First of all thank you for developing such a great product! Already using it for relational sources and it works flawlessly. I am now trying to load data to BigQuery using Mailchimp source. Connector setup works with no problem, however the sync process always ends up with 0 records. I hope i am not missing something trivial here, but I cannot find any hint on why no record is being read.

Aldo, is possible to share the sync log? Maybe there something there can help

Hi @[DEPRECATED] Marcos Marx, I found the possible cause in the log:

2021-06-15 14:51:06 ERROR (/tmp/workspace/8/0) LineGobbler(voidCall):85 - requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: <https://us2.api.mailchimp.com/3.0/campaigns?count=100&sort_field=create_time&sort_dir=ASC>

Since Airbyte UI reported the job as

Succeeded

I thought there was some other misconfiguration. As an additional info, Mailchimp API needs two parameters, a datacenter

dc

and an

apikey

, as seen here. Airbyte’s Mailchimp source configuration only asks for an

username

and the

apikey

. The

dc

parameter is used in Mailchimp to build the URI of the API server (like

<https://us2.api.mailchimp.com/3.0/campaigns>

). However, in our case the

dc

parameter should be

us10

and NOT

us2

, like it appears in the error log. So it looks like we need to find a mechanism to instruct Airbyte Mailchimp source connector to use the

dc

parameter correctly to build the API URI

@Aldo Chiecchia could you open a Github issue about this? Did you follow the Getting started steps to generate your APIKEY?

@[DEPRECATED] Marcos Marx I already had an APIKEY, so I will try to follow exactly your recommendation first as suggested in “Getting started”. If the issue is not solved I will open an issue.

Not the OP, but we are trying to use the same

MailChimp

connector and are running in to a similar issue. We did some digging in the connector code and noticed in the streams that

url_base

is being initiated as

"<https://us2.api.mailchimp.com/3.0/>"

. This isn't ideal, since each url_base is different and is based on the api_key we get from mailchimp (ours is us13, as an example). We could change that for each api key or write

url_base()

functions for each stream but we also noticed in the

mailchimpclient.py

file has a base_url that is actually created from the api key. What we can't seem to figure out is how this is being used in the process, as it seems that without

url_base

it can't run any of the streams at all. We can hardcode the datacenter change for the

url_base

in streams and all of the checks and behavior work as expected. I'd be happy to open a GitHub issue as well if that works better or if one hasn't already been put together. Lastly, I want to say that it has been a pleasure using Airbyte and it has been awesome to work with. I love watching it grow every day.

@Steven Anderson the ideal is have an datacenter parameter to choose, correct?